Homing endonucleases are extraordinarily specific DNA-binding proteins, acting specifically at individual sites within a host genome. These proteins are under instense study for the purpose of engineering single chain gene-specific reagents to be used for gene therapy and other applications. Over the past 10 years, we have determined the structure and mechanisms of representatives form all known families of homing endonucleases, found respectively in phage, eubacteria, archae, and single cell eukarya. In addition, we have described the creation of homing endonuclease variants that act at noncognate sites. These constructs have been generated using both bacterial selection strategies and compuational methods, both of which target enzyme residues that directly contact DMA basepairs. In either case, such experiments have produced endonucleases that display shifted DMA recognition properties, but at the cost of reduced site-discrimination abilities. We hypothesize that in order to completely reprogram the DNA recognition specificity of a homing endonuclease, without a reduction in site discrimination, the resculpting of protein-DNA contacts must be combined with the selection of structural mutations in the nearby enzyme scaffold that "fine-tune" the protein -DNA interaface of each novel cognate complex. The goal of overall Specific Aim 1 of the Northwest Genome Engineering Consortium is to accomplish this task by combining somatic hypermutation of the endonuclease scaffold, computational redesign and selection of DNA contacts, and biochemical/biophysical characterization of the resulting endonuclease constructs. In our component of the consortium's activities, we will be responsible for the following aims: 1. We will determine the in vitro site specificity profile of the novel endonuclease construcst using two related methods to directly visualize cleavage of DNA target variants and to quantitate specificity at each base pair. 2. We will determine the thermodynamic signature of cognate and non-cognate site recognition for redesigned homing endonucleases, using isothermal titration calorimetry (ITC). 3. We will determine the three-dimensional structure of novel endonuclease-DNA cognate pairs at high resolution, and will characterize (a) the effect of enzyme scaffold mutations on backbone structure, and 9b) the accuracy of computational redesign predictions within the protein-DNA interface.